**** BEGIN LOGGING AT Tue Dec 10 02:59:58 2013 Dec 10 09:29:34 good morning Dec 10 10:30:57 morning all Dec 10 10:35:52 gm bluelightning, all Dec 10 10:56:14 How Do I limit maximum number of simultaneous SSH sessions to 3 Dec 10 10:56:26 cat /etc/issue Dec 10 10:56:46 Poky 9.0 (Yocto Project 1.4 Reference Distro) 1.4.1 Dec 10 10:57:34 I don't have /etc/security/limits.conf on my box with image compiled through yocto Dec 10 10:58:03 ramose: if you're using openssh, then you can read the openssh manual Dec 10 10:58:14 the sshd has an option for that Dec 10 11:15:32 rburton: already tried these options MaxSessions 3 in /etc/ssh/sshd_config but didn't work out. Dec 10 11:16:07 Do i really need PAM module for that Dec 10 11:16:07 ? Dec 10 11:19:39 ramose: no idea, i guess you should speak to the openssh maintainers Dec 10 11:23:52 ok, fine rburton Dec 10 11:24:06 <_tid> "Specifies the maximum number of open sessions permitted per network connection" i guess if you reconnect using another network connection this limit does not apply Dec 10 11:27:02 <_tid> the feature you are looking for seems to be in pam not in the sshdconfig Dec 10 11:30:19 the recent changes to bitbake's git fetcher wrt verifying the presence of a sha1 is causing some issues when using git tags with the url tag= parameter. the tag is resolved to a sha1 of the tag object, not the commit and bitbake errors out with it not being able to find that sha1 on the branch. Dec 10 11:34:52 any ideas for fixing this issue? Dec 10 11:36:32 RP: ^ Dec 10 11:39:26 zibri: does specifying the branch the tag is on help? Dec 10 11:40:23 rp: nope Dec 10 12:18:31 zibri: I'll need to setup a test case and figure out how we can handle it then... Dec 10 12:18:57 right. i was hacking around, but wasn't able to find a nice solution. :( Dec 10 12:19:46 i did try to write some initial unittests for the git fetcher; something you'd be interested in? can send to the list after some final touches Dec 10 12:21:34 zibri: there are already some fetcher unittests? Dec 10 12:21:44 zibri: extra ones are welcome though Dec 10 12:21:56 yes, but only for the base class, not the git fetcher specifically, right? Dec 10 12:23:06 zibri: well, there are git urls in those tests Dec 10 12:23:07 there are some high level tests that do test the git fetcher. Dec 10 12:23:29 zibri: right Dec 10 12:23:53 i can send them and you can see if they are worth merging Dec 10 12:24:36 zibri: sure Dec 10 12:35:03 _tid : "Specifies the maximum number of open sessions permitted per network connection" i guess if you reconnect using another network connection this limit does not apply Dec 10 12:35:14 would you explain your point a bit Dec 10 12:56:13 zibri: I can see what you mean looking at the code, this is tricky :/ Dec 10 13:38:23 zibri: http://git.yoctoproject.org/cgit.cgi/poky-contrib/commit/?h=rpurdie/t2&id=d211662ca92ee084c0c4a3d343e0b1f28ab853ac Dec 10 14:11:42 What is the update command on yocto build? apt-get or rpm or opkg? Dec 10 14:18:29 rp: it seems to be working, great! thanks a lot Dec 10 14:26:24 parekh: depends on what package manager you specified when building the images Dec 10 14:26:38 parekh: default is rpm, iirc Dec 10 14:35:48 parekh: if you're using rpm the tool we use around that for downloading updates / resolving dependencies is called "smart" Dec 10 14:47:03 rburton: I used the default Images from gumstix website. Dec 10 14:57:22 parekh: I'm guessing that would be using opkg then Dec 10 14:58:55 bluelightning: Ok. Will try that. Dec 10 15:53:37 jwessel: please update https://bugzilla.yoctoproject.org/show_bug.cgi?id=4100, thanks Dec 10 15:53:38 Bug 4100: enhancement, Medium+, 1.6, jason.wessel, RESOLVED FIXED, Add EFI support to ISO images Dec 10 15:59:22 hi, what is the best way to "downgrade" a patch made agaisnt the latest linux staging kernel version? Dec 10 15:59:32 e.g. using the denzil, dylan, etc kernels. Dec 10 16:00:07 he Yocto Project Tech Meeting Con-Call will be starting at the top of the hour Dec 10 16:00:07 Dial-in number: 1.972.995.7777 / Participant passcode: 42001078 Dec 10 16:00:07 This call is open to all and the channel remains open to discuss any topic Dec 10 16:00:37 YPTM: good morning Dec 10 16:00:43 YPTM: Saul is on Dec 10 16:00:49 YPTM: Matthew is on Dec 10 16:01:00 YPTM: Scott Rifenbark has joined the call. Dec 10 16:01:05 YPTM: Polk is on Dec 10 16:01:13 YPTM: Bruce is on Dec 10 16:01:17 YPTM: Nitin is on the call Dec 10 16:01:26 YPTM: welcome to the technical team meeting, please let me know who's on the bridge. Thanks Dec 10 16:01:31 YPTM: ross on Dec 10 16:01:39 YPTM: LaurentiuP joined Dec 10 16:01:43 * rburton wonders what music to put on today Dec 10 16:01:46 YPTM: Niklas on Dec 10 16:02:09 YPTM: Paul Eggleton is on Dec 10 16:02:11 YPTM: AlexG here too Dec 10 16:02:27 YPTM: Beth Flanagan & Alex Damian are also in the room here Dec 10 16:02:29 * RP is on the call Dec 10 16:02:33 YPTM: cristian is present Dec 10 16:02:53 YPTM: Corneliu joined Dec 10 16:02:55 YPTM: current QA status for 1.6_M1.rc1: https://wiki.yoctoproject.org/wiki/1.6_QA_Status Dec 10 16:03:04 YPTM: Tom Z here Dec 10 16:03:25 Song_Liu: YPTM: Richard is on the call Dec 10 16:03:41 YPTM: any open? Dec 10 16:03:53 YPTM: jzhang on the call Dec 10 16:04:04 Song_Liu: YPTM: 1.5.1 planning Dec 10 16:04:16 sgw_: also 1.4.3 Dec 10 16:04:27 er Song_Liu: also 1.4.3 Dec 10 16:04:29 Song_Liu: ^^^^ Dec 10 16:06:42 YPTM: Page Song_Liu is refering to: https://wiki.yoctoproject.org/wiki/Yocto_Project_v1.6_Status Dec 10 16:08:20 YPTM: Michael joined Dec 10 16:14:48 YPTM: Björn joined Dec 10 16:15:55 Song2: must drop the call in 5 minutes Dec 10 16:23:39 Valentin, are you here? Dec 10 16:25:15 FYI all: I am moving jobs. Valentin Corbea will take over the ptest lead role at enea starting 2014. Dec 10 16:25:49 Zagor: vanentin will have big shoes to fill Dec 10 16:26:02 rburton: :-) thanks Dec 10 16:26:27 Zagor: thanks for all of your work, and the heads up - will you be working with Yocto/OE in your new job at all (if I may ask)? Dec 10 16:27:33 YPTM: thank you all for joining the meeting. Have a nice day/evening Dec 10 16:27:33 Zagor: You'll be missed, I hope you continue to be involved in the project somehow! Dec 10 16:27:39 bluelightning: not like today, no. possibly as user. Dec 10 16:30:20 Zagor: Just saw your message. Yes, thank you for your support in the past years. Good luck to your new job Dec 10 16:31:26 I'll be at work (and in this channel) until dec 20. Dec 10 16:39:33 Hi. I'm trying to add AUFS for Freescale MX6 (kernel 3.10). Does anyone know if this is supported? Dec 10 16:43:38 nikmol: if the kernel you are using is linux-yocto, the AUFS patches should be in the kernel source already; if not, you may be on your own Dec 10 16:44:02 has anyone seen my question before the meeting? (is it still ongoing?) Dec 10 16:45:24 bluelightning: I'm using the meta one for Freescale. If the paths is there, do I have to enable it or is it on by default? Dec 10 16:46:26 nikmol: no idea I'm afraid, I don't deal much with FSL BSPs ; otavio is probably the best person to ask about that Dec 10 16:46:43 it looks like it doesn't use linux-yocto, FWIW Dec 10 16:46:53 lpapp: the meeting is done Dec 10 16:47:11 lpapp: FWIW, I didn't understand your question Dec 10 16:48:01 bluelightning: why not? Dec 10 16:48:26 lpapp: er, because I don't know what you mean by "downgrade a patch"... Dec 10 16:48:48 backport? Dec 10 16:48:59 backport, yes. Dec 10 16:49:59 lpapp: well, your question is slightly similar to "how long is a piece of string" Dec 10 16:50:08 lpapp: i.e, it depends on the patch Dec 10 16:50:48 no, I am asking a generic patch backport question with Yocto. Dec 10 16:51:02 obviously, I could grab the same tarball outside, but it would be nice if this could all be done within Yocto. Dec 10 16:51:31 because I develop new features agains the latest upstream for sure... I mean, if I wanna upstream them, I have to. Dec 10 16:51:43 but on the other hand, we use an older version within Yocto (enforced by Yocto) Dec 10 16:53:09 lpapp: you can use the git repos that yocto cloned for you if you want to avoid cloning again Dec 10 16:53:53 bluelightning: thanks, should I still check with otavio to make sure? Dec 10 16:54:00 rburton: it is a tarball within Yocto. Dec 10 16:54:07 e.g. denzil had 3.2 iirc. Dec 10 16:54:59 nikmol: well hopefully just mentioning his name should prompt him to reply when he is around, but you can try pinging him later if not Dec 10 16:55:24 I guess Yocto fails at applying the change. Dec 10 16:55:33 what are the steps then to update the patch and resolve the conflict? Dec 10 16:55:54 without using an external decompression and workflow outside Yocto, presumably dealing with the downloaded and decompressed tarball. Dec 10 16:56:14 bluelightning: He already send me an e-mail. So I have my answer. Thanks Dec 10 16:56:21 lpapp: if you're talking specifically about a kernel then linux-yocto is built from git repos Dec 10 16:56:22 nikmol: ah ok, great Dec 10 16:56:23 which will help Dec 10 16:56:44 lpapp: apply patch, see what breaks, apply by hand in a text editor. Dec 10 16:56:44 rburton: strange, but ok. Dec 10 16:57:40 that seems like the standard way to backport a patch. Dec 10 16:57:42 i find meld useful for manual backports of patches, as you can see three forms of the same source at once with diffs Dec 10 16:57:45 BSDCat: pretty much Dec 10 16:58:28 yocto doesn't (and can't!) help backports any more than giving you access to the source tree so you don't have to re-fetch it Dec 10 16:59:44 rburton: actually, it could and should IMO Dec 10 16:59:56 lpapp: explain how then Dec 10 17:00:03 failing to apply a patch, it could hint or even open the conflict file. Dec 10 17:00:17 once that is done, you could have a command for patch regeneration with bitbake Dec 10 17:00:26 which overwrites the original patch file. Dec 10 17:00:38 so the only thing you would need is the actual resolution, and not the boilerplate around. Dec 10 17:00:43 doesn't patch.bbclass do that? Dec 10 17:01:05 not as far as I am ware. Dec 10 17:01:06 a Dec 10 17:01:15 yeah, UserResolver Dec 10 17:01:20 not sure how to activate it, never used it. Dec 10 17:01:20 it shows what file conflicts, but it does not open up with conflicting line markers. Dec 10 17:01:29 and once I resolve it manually, it is up to me to regenerate Dec 10 17:01:48 try setting PATCHRESOLVE="user" Dec 10 17:05:10 bluelightning: Since I'm pretty new to this community, I just wanted to get a more understanding of how this works. You mentioned linux-yocto kernel. Does each BSP project use a different linux-kernel? Dec 10 17:06:42 nikmol: well, some do, some don't; our recommended practice is to base your BSP's kernel recipe on linux-yocto so that you have the benefit of the config fragment and git branch handling that it provides, but some prefer to do their own thing instead Dec 10 17:07:53 bluelightning: thanks. Dec 10 17:08:21 rburton: will that also regenerate the patch automatically if I say, I finished the resolution? Dec 10 17:10:35 lpapp: possibly. Dec 10 17:10:51 lpapp: looks like it *can* Dec 10 17:11:03 depending on what patch set implementation was used. Dec 10 17:11:06 worth a go! Dec 10 17:42:29 Hi, Is there a way to force some packages to build non parallell? I know that something is wrong when one have to force such things, but just having it as a work around until the root problem is found? Can this be done in a bb or bbappend? Thanks! Dec 10 17:44:27 assa: it can indeed, just do PARALLEL_MAKE = "" in the bb/bbappend Dec 10 17:44:55 assa: is this a recipe that we provide or one you've created? Dec 10 17:45:47 bluelightning: something I've created, so you can just stay calm ;-) Dec 10 17:45:55 Thanks for your help! :) Dec 10 17:45:59 assa: ok, just checking :) Dec 10 17:46:01 assa: np Dec 10 17:53:18 rburton: what am I supposed to do after a successful quilt refresh? Dec 10 17:53:36 it does not put me back to the build directory.... it leaves me in the linux folder... that does not quite look right... Dec 10 17:53:51 shouldn't it put me back to the build folder where I was? Dec 10 17:54:38 moin Dec 10 17:54:43 hi Dec 10 17:55:09 oh, I had to press ctrl-d Dec 10 17:55:30 if only I had received any hint ... ;-) Dec 10 17:55:44 bluelightning: looks like you got your wish Dec 10 17:56:08 mr_science: which one, sorry? Dec 10 17:56:14 mr_science: not the pony, I'm guessing Dec 10 17:56:15 ;) Dec 10 17:56:35 not unless nginx comes with a a pony... Dec 10 17:57:00 ah right Dec 10 17:57:02 :) Dec 10 17:57:40 i got sidetracked after the second submission but i noticed some activity on the ml Dec 10 17:58:07 seems like i owe someone a pitcher and a cookie... Dec 10 18:01:12 something must have gone wrong with the manual patch conflict resolution as I clearly have syntax errors in my code, yet Yocto builds it fine.... or it does not build it for some reason. Dec 10 18:04:06 right, defconfig was not updated...... Dec 10 18:04:21 why not? Shouldn't Yocto update that for bitbake virtual/kernel -c menuconfig? Dec 10 18:05:32 so is there a Yocto config option for updating the defconfig in files? Dec 10 18:05:44 rburton: ^ Dec 10 18:05:57 hi there Dec 10 18:06:02 I have a patch for minicom recipe Dec 10 18:06:13 is that alright if I send it using mail? Dec 10 18:06:29 Xz: yes Dec 10 18:06:52 to: openembedded-core@lists.openembedded.org Dec 10 18:06:59 with patch as a body of message Dec 10 18:07:42 Xz: yep Dec 10 18:08:14 done Dec 10 18:09:04 RP: maybe you know this defconfig update question? Dec 10 18:12:17 hi Dec 10 18:12:20 https://github.com/woglinde/meta-java/issues/49 Dec 10 18:12:26 Xz: email is how we usually submit changes, yes. best to use git send-email / git format-patch to avoid mail client's screwing with things, though Dec 10 18:12:30 maybe someone has an idea? Dec 10 19:32:56 otavio: you around and aware of the fetch issues for meta-fsl-ppc? Not sure who I should forward those to? Dec 10 20:41:52 sgw_: I am around Dec 10 20:42:14 sgw_: it ought to forwarded to Luo Dec 10 21:55:11 otavio: thanks Dec 10 22:14:42 <_tid> is there any way to put kernel modules in the generated .ubi ? Dec 10 23:03:23 sgw_: yw **** ENDING LOGGING AT Wed Dec 11 02:59:59 2013